Differentiation and Cell Growth by Symmetrical and Asymmetrical Mitosis: A Hypothesis

DIFFERENTIATION AND CELL GROWTH BY SYMMETRICAL AND ASYMMETRICAL MITOSIS: A HYPOTHESIS H. W. v. HEYDEN and DORIT v. HEYDEN* Introduction The human bone marrow produces daily up to 500 grams of blood cells of all kinds, including approximately 100 grams of granulocytes [1, 2]. Essentially a self-producing phase coexists within a differentiation system. We have used myelopoiesis as an example of a differentiation system. Cell Production in a Differentiation Unit The basis of any differentiation system is mitosis. One dividing stem cell, reproducing two identical daughter cells, supposedly allows only self-reproduction but no differentiation; this kind of mitosis is termed symmetrical reproduction mitosis (sR). It is analogous, but based on terms of differentiation, to a symmetrical nondifferentiation mitosis (snD) (figure la). It is necessary in a differentiating system that the mitosis of stem cells provide one cell for self-reproduction and one cell for differentiation [3, 4]. This kind of stem cell mitosis we term semisymmetrical reproduction mitosis (ssR) or semisymmetrical differentiation mitosis (ssD) (fig. 1/3). There is also the possibility of a symmetrical differentiation mitosis (sD) or symmetrical nonreproduction mitosis (snR) (fig. Ic). Here, the two daughter cells are identical to each other but do not have the characteristics of the mother cell. * Medical University Klinik II, Otfried Mueller Street, 74 Tübingen, Federal Republic of Germany. This work was supported by the Mary B. and L. H. Marshall Foundation and funds from Deutsche Forschungsgemeinschaft. We wish to thank Dr. G. E. Moore and Dr. W. Rhomberg for their discussion and encouragement and also Mr. Ronald P. Koch and Miss Nancy J. Stuhlmiller for their assistance in the preparation of this manuscript . 348 I H. W. v. Heyden and Dorit v. Heyden · Differentiation and Cell Growth T T© Fie. 1.—The different kinds of mitosis: ®= stem cell; (D= first differentiated cell = myeloblast, a, Symmetrical reproduction mitosis (sR) = symmetrical nondifferendation mitosis (snD). b, Semisymmetrical differentiation mitosis (ssD) = semisymmetrical reproduction mitosis (ssR). c, Symmetrical differentiation mitosis (sD) = symmetrical nonreproduction mitosis (snR). In myelopoiesis, the myeloblast is the first differentiated cell. Of the three mentioned mitosis possibilities, which one is involved in the case of the myeloblast cell? Theoretically, there are three possible derivatives from one myeloblast: two myeloblasts (analogous to fig. Ia); one myeloblast and one promyelocyte, the second differentiated cell (analogous to fig. Ib); or two promyelocytes (analogous to fig. Ic). One further possibility is that the myeloblast would no longer divide, which we term maturation. The only strong probability is that one myeloblast divides into two promyelocytes by sD, based on the relative percentage distribution of cells within the myelocytic series [5-8]. In figure 2, we simulated a ratio of one myeloblast to two promyelocytes to four myelocytes to eight metamyelocytes to eight neutrophils. If one myeloblast proliferated on the basis of one sR, we would have only myeloblasts, as in the case of acute myelogenous leukemia·»* ·:· < O©® *?·ß· ? O ? ?0T FiG. 2.—Cell pattern in a differentiation unit: ®= stem cell; f= first differentiated cell = myeloblast;©= second differentiated cell = promyelocyte;®= third differentiated cell = myelocyte; f= fourth differentiated cell = metamyelocyte;®= fifth differentiated cell = neutrophil, a = semisymmetrical differentiation mitosis; b — symmetrical differentiation mitosis; e = maturation process. The determination lane is demonstrated by®—f—@-~®—f—® Perspectives in Biology and Medicine · Spring 1973 | 349 (AML). If a myeloblast divided by one ssD, we would have less neutrophils in comparison with other differentiated cells as is the case in chronic myelogenous leukemia (CML). Maturation from one myeloblast to one neutrophil can be excluded since this would result in a one-to-one ratio between cells of the granulocyte series. We must also consider the mitosis possibilities for the second, third (myelocyte), and fourth (metamyelocyte) differentiated cells. To obtain the minimal ratio of one myeloblast to eight neutrophils, it is necessary that promyelocytes and myelocytes divide by sD too. As demonstrated in figure 2, our model of one differentiation unit is described by seven sD and, finally, eight maturation processes. One differentiation unit consists of one myeloblast, two promyelocytes, four myelocytes, eight metamyelocytes, and eight neutrophils. One myeloblast has the potential of producing eight neutrophils. Such a differentiation unit allows only a certain number and proportion of differentiated cells. This could mean that the potential for logarithmic growth is limited to that number of cells which the consecutive differentiation mitoses, including the final maturation processes, allow. A differentiation unit is, therefore, an example of self-proportioning and self-limiting cell production. A differentiation unit should be a regulated system. In the case of inflammation, for example, the capacity to produce neutrophils can be reached if the turnover rate within this differentiation unit cannot be accelerated. Cell Production in CML and AML On a cellular basis, most of the leukemias in human peripheral blood and bone marrow are expressed in an abnormally high cell concentration and an abnormal cell distribution. Patients with CML have all kinds of cells in the granulocyte series, whereas patients with AML have predominantly one kind, either myeloblasts or promyelocytes. The cellular situation of CML is illustrated in figure 3, a-c, where one myeloblast divides by one ssD (3a), one promyelocyte divides by one ssD (3fr), or one myeloblast and one promyelocyte divide by ssD (3c). In each case, the myeloblasts and promyelocytes have stem-cell characteristics. This means that the differentiation unit has partially lost its ability to control the cell number. 350 1 H. W. v. Heyden and Dorit v. Heyden · Differentiation and Cell Growth ? ? ? O ©© T©_? O T ? ? ? 9 O O O T ? ? ? O©© ?© ? T O O ?© ? ? ?© ? T T O T ? T ? ?© T© O O ? T O O T T ? T ? O Fie. 3.—a, b, c, Cell pattern in chronic myelogenous leukemia; d, e, cell pattern in acute myelogenous leukemia. a,0divides by ssD. fe,(2)divides by ssD. c,®+@divide by ssD. d,®divides by sR. e,@divides by sR. In terms of the clonal theory, two existing populations are derived from two stem cells, one malignant and one normal. In comparison, the suggestion is made here that all cells may be derived from one stem cell (fig. 3, a-c). As diagrammed in figure 3, d and e, AML and acute promyelocytic leukemia are the result of sR (snD) which does not produce any further differentiated cells. In comparison to the myeloblast or promyelocyte in CML, each cell has double stem-cell characteristics. In AML there is no possibility of proportioning and self-limitation. Perhaps in AML exists the potential of exponential cell growth. This model of differentiation explains the nonexistence of acute myelocytic leukemia (the third differentiated cell). A metamyelocyte is more differentiated than a myelocyte, the latter more than a promyelocyte , the latter more than a myeloblast, and the latter more than a stem cell. A more differentiated cell has less stem-cell potency than a less differentiated cell, that is, the only mitotic possibility of differentiated normal cells is the sD, and the metamyelocyte has only the possibility of maturation. In situations of leukemic cell growth, only the first and second differentiated cell still has enough stem-cell potency for a snD or a ssD. Differentiation means, therefore, that, along the determination lane (fig. 2), an increase in differentiation parallels a decrease in stem-cell characteristics. A myelocyte seems to be too differentiated for a snD or ssD. Perspectives in Biology and Medicine · Spring 1973 | 351 Discussion It has been the aim of this paper to describe the possibility of cellpopulation regulation by differentiation and the consequence of cell growth in CML and AML if differentiation does not exist, either partially or completely. As an example, we examined the balanced situation of a differentiation unit found in myelopoiesis and the unbalanced leukemic situation. In the cellular terms used here, there is no difference between AML and the terminal blastic phase of CML. The similarity was recently discussed from the viewpoint of chromosomal pattern [9] and cell kinetics [10]. The clonal theory is usually defined as two existing cell populations derived from one malignant and one normal cell. Based on our model (fig. 3, a-c), the ssD gives a similar pattern of cell distribution to that which would exist if two parallel existing populations were derived from two cells. This does not exclude the existence in CML and AML of normal myeloblasts dividing by sD. Cell production in CML stands between the normal cell production of the differentiation unit and the cell production of AML. As figure 3, a-c, demonstrates, part of the cells produced in CML are neutrophils at any given time of observation. Killmann also discussed the possibility that normal-appearing neutrophils were derived from malignant stem cells [H]. Cell production in a differentiation unit is not only self-proportioning and self-limiting but also has some capacity to accelerate the turnover rate, perhaps in response to demand [12]. If self-proportioning and self-limitation of cell number by differentiation does not exist, the cell production could best be compared to that of an exponential growing system at the very onset of AML. The cell situation in untreated AML and in a very advanced stage of disease could be said to be in the stationary or semistationary phase. Absence of differentiation does not mean that multiplication of cells cannot be regulated. For easier understanding, let us assume in the case of acute leukemia a closed system, open only for plasma, serum, or nutritional medium, with one myeloblast dividing by snD at time 0. The cell growth will at first be exponential, but, with increase of time, limitations force the system of dividing cells to adaptation. The log phase (growth in logarithmic fashion) is followed by a semistationary phase 352 I H. W. v. Heyden and Dorit v. Heyden · Differentiation and Cell Growth (probably described by a Gompertzian function [13, 14]) which, in turn, ends in a stationary phase—a cell-growth system expanding no further. The external forces, as limiting factors for the potential exponential growth of cells in acute leukemia, would be high cell density , presence of inhibition factors, or perhaps exhaustion of nutrients . The expression of cell population being forced to adaptation could be prolongation of cell cycle, an increase of cells resting in Gi phase, respectively, being out of circle as G0 cells, a decreased growth fraction, an increase of doubling time, and increase in the rate of cell death. Some of these parameters of cells dividing by snD in the logarithmic, semistationary, and stationary phases were recently described by a model for acute leukemia in vitro [15]. With exceptions, these comparative parameters are similar for acute leukemia in vivo [13, 15-17]. One of these exceptions is that some undifferentiated cells in acute leukemia are able to become differentiated [15, 16]. In terms used here, we may say that a certain fraction of cells in AML oscillate between snD, ssD, and sD. Remission would mean that the ssD or snD is "switched" to the advantage of sD. The existence of normal myeloblasts dividing by sD is, as mentioned above, not excluded. If we adopted the same closed system for one normal myeloblast dividing by sD, we would observe the formation of one differentiation unit and nothing more. Therefore, differentiation is that force which keeps the system from being filled with cells. The cell production in a differentiation unit is only possible in one direction (in the sense of a vector), whereas the cell production in AML has no direction. Normal cells in the steady state (cell production = cell loss) do not have a doubling time [13] as is true for expanding leukemic cell populations [13]. For description of cell production of normal cells, it would be preferable to use the term "turnover rate" [18]. The cell production in CML stands between the cell production of the described differentiation unit and AML; we assume one or two different cell types dividing by ssD (fig. 3, a-c). The differentiation unit demands a steady influx of normal myeloblasts . For cells dividing by snD, as in the case of acute leukemia, this is not necessary; these cell populations are self-maintaining [16]. This is in agreement with data obtained by Clarkson et al. [16] but is in some conflict with the stem-cell hypothesis for leukemic cells [19]. The concept of asymmetrical or symmetrical mitosis of blood cells Perspectives in Biology and Medicine · Spring 1973 | 353 was, to our knowledge, first given in several papers by Osgood [3, 4]. This concept is then mentioned in the textbook of Wintrobe's Clinical Hematology [20]. Also, Mendelsohn and Takahashi [21] and Clarkson and Fried [22] mentioned the symmetry or asymmetry of mitosis, the latter ones referring to Lajtha [23] and Osgood [24]. Asymmetrical division could be thought of as a concept of cells during embryogenesis [25-27]. Several eggs (e.g., amphibia eggs) demonstrate an asymmetrical distribution of presumptive regions. Within the same region, cell mitosis does not seem to be symmetrical because the quantity of yolk does not seem to be equally distributed within progeny cells of the same generation. Asymmetrical distribution of any quality or quantity within one cell complex or one cell can be due to "asymmetrical acting external forces" [28] such as gravity or nutritional gradients [25-28]. What could be explained for fertilized eggs is much more difficult to explain for normal stem cells. To give only one reason, the identification of normal stem cells is very difficult and only approaches seem tobe possible [12, 29, 30]. For the normal stem cell we assume, therefore, asymmetrical arrangement of receptors for inducers, inhibitors, and repressors. Chalones [31] or the colony-forming factor [32-34] could be regulating or differentiating effectors. How could the value of the given hypothesis be tested? If asymmetrical divisions occur in different kinds of tissues, the cells themselves must either be asymmetrical or polar as is assumed for the stem cells in the hematopoietic tissues. Polar cells are thought to be present in any regenerating tissues. Theoretically, there are several possibilities for asymmetrical distribution during mitosis. Either the membrane or the plasma, or a combination of them could be unequally distributed to the daughter cells. In polar cells, the mitotic spindle should have a distinct and fixed position in relation to the membrane surface. The relationship of the Ph1 chromosome in CML cells to asymmetrical ssD must be determined. The Ph chromosome may be observed in CML, as demonstrated by figure 3, exclusively in the myeloblasts (fig. 3a), in the promyelocytes (fig. 3b), and in both (fig. 3c). If a translocation of the missing parts to another chromosome within the same cell could be excluded [9], one might assume on the basis of the given hypothesis that the missing parts are located in the nor354 ] H. W. v. Heyden and Dorit v. Heyden · Differentiation and Cell Growth mal-appearing daughter cells (fig. 3, a-c). The next question would be, What is the fate of the missing part in the normal-appearing cells dividing by sD? If these defective chromosomes are not digested as corpora aliena, one could use these part chromosomes as markers in order to trace an asymmetrical mitosis. We propose the existence of symmetrical and asymmetrical mitosis at the cellular level using myelopoiesis as an example. Both at the molecular and the macromolecular level, there exist many examples for symmetry and asymmetry. We refer only to a few of them: the symmetry and asymmetry of the carbon atoms, eis- and transisomers and 1- and d-amino acids. It still remains an open question as to why only 1-amino acids are required for protein synthesis. Repetitious sequences of DNA strands are strict expressions of symmetry at the macromolecular level. Symmetry at the macromolecular level was recently discussed as a function of different biological systems [35]. REFERENCES 1.G. E. Cartwright, J. W. Athens, and M. M. Wintrobe. Blood, 24:780, 1964. 2.S. Perry. New Eng. J. Med., 285:214, 1971. 3.E. E. Osgood, J. Nat. Cancer Inst., 18:155, 1957. 4. ----------. Perspect. Biol. Med., 6:228, 1963. 5.P. K. Lala, M. A. Maloney, and H. M. Patt. Acta Haemat. (Basel) , 31 : 1 , 1964. 6.S. A. Killman, E. P. Cronkite, T. M. Fliedner, V. P. Bond, and G. Brecher. Blood, 21:141, 1963. 7.J. LoBue. In: A. S. Gordon (ed.). Regulation of hematopoiesis, p. 1167. New York: Appleton-Century-Crofts, 1970. 8.M. M. Wintrobe (ed.). Clinical hematology, p. 37. Philadelphia: Lea & Febiger, 1967. 9.A. A. Sandberg and D. K. Hossfeld. Ann. Rev. Med., 21:379, 1970. 10.H. A. Godwin, T. S. Zimmerman, and S. Perry. Blood, 31:686, 1968. 11.S. A. Killmann. Ser. Haemat., 1:103, 1968. 12.E. P. Cronkite. New Eng. J. Med. 282:683, 1970. 13.H. E. Skipper and S. Perry. Cancer Res., 30:1883, 1970. 14.V. T. Devita. Cancer Chemother. Rep., 2:23, 1971. 15.A. Todo, A. Strife, J. Fried, and B. D. Clarkson. Cancer Res., 31:1330, 1971. 16.B. Clarkson, J. Fried, A. Strife, Y. Sakai, K. Ota, and T. Okhita. Cancer, 25:1237, 1970. 17.B. D. Clarkson and J. Fried. Med. Clin. N. Amer., 55:561, 1971. 18.M. Lipkin. In: R. Baserga (ed.). The cell cycle and cancer, p. 13. New York: Dekker, 1971. 19..F. Gavosto and A. Pileri. In: R. Baserga (ed.). The cell cycle and cancer, p. 97. New York: Dekker, 1971. 20.M. M. Wintrobe (ed.). Clinical hematology, p. 18. Philadelphia: Lea & Febiger, 1967. 21.M. L. Mendelsohn and M. Takahashi. In: R. Baserga (ed.). The cell cycle and cancer, p. 68. New York: Dekker, 1971. Perspectives in Biology and Medicine · Spring1973 | 355 22.B. D. Clarkson and J. Fried. Med. Clin. N. Amer., 55:576, 1971. 23.L. G. Lajtha. In: L. O.Jacobson and M. Doyle (eds.). Erythropoiesis, p. 140. New York: Grune & Stratton, 1962. 24.E. E. Osgood. Geriatrics, 19:208, 1964. 25.W. S. Bullough. The evolution ofdifferentiation. London: Academic Press, 1967. 26.N. T. Spratt, Jr. Introduction to cell differentiation. New York: Reinhold, 1964. 27.C. A. Pasternak. Biochemistry of differentiation. New York: Wiley, 1970. 28.J. D. Watson. Molecular biology of the gene, p. 51 1 . New York: W. A. Benjamin, 1970. 29.S. Perry. Ann. Rev. Med., 37:171, 1971. 30.D. W. van Bekkum, M.J. van Noord, B. Maat, and K. A. Dicke. Blood, 38:547, 1971. 31.W. S. Bullough and T. Rytomaa, Nature, 205:573, 1965. 32.M. Paran, L. Sachs, Y. Barak, and P. Resnitzky. Proc. Nat. Acad. Sci. USA, 67:1542, 1970. 33.P. A. Chervenick and D. R. Boggs. Blood, 37:131, 1971. 34.W. A. Robinson, J. E. Kurnick, and B. L. Pike. Blood, 38:500, 1971. 35.J. Monod. In: A. Engstrom and B. Strandberg (eds.). Symmetry and function of biological systems at the macromolecular level, p. 15. New York: Wiley, 1968. HYMN TO A LONELY ALCOHOLIC Glazed eyes float in an unknown face Only briefly blind to lonely, weaving on a loom of languid laughter. Garbled curses, shroud his staggering form casting crooked shadows backward. A dimly recorded youth in reckless pursuit of pleasure come easy: His only mistress. Now old age empty his master. Come, some distant and relentless gutter, sing one last hymn to lonely. Peter A. Olsson, M.D. 356 [ H. W. v. Heyden and Dorit v. Heyden · Differentiation and Cell Growth ...